Josh Hopkins has a thoughtful article over at The Space Review on the issues that must be addressed by proponents of propellant depots. I may have a response a little later, but I would note that Boeing has done a lot of work on the concept, and may have answers to some of the questions. In any event, we could have resolved them with a tiny fraction of the money that we just pissed away on Ares over the past four years.
[Update early afternoon]
In comments over there, Jon Goff makes one of the points that I would have had I gotten around to it (we’re getting ready to move, and writing SBIRs among other things, around here). It’s worth repeating:
2-Regarding launch costs: I think most depot researchers, like myself, would agree that if the cost for propellant delivery were truly likely to be in the $30-60kg range, that depots weren’t likely to make sense. However, is using delivery of cargo to a manned space station, where the cargo carriers themselves have to function as pressurized space station modules, really that realistic of a starting point for estimating the cost of propellant delivery to a likely unmanned depot? I know that that was probably the easiest way to do the analysis, since those numbers are available…but aren’t we setting up a strawman here? First you pick the obviously most expensive route (having the tankers be fully-functional prox-ops vehicles), and then pick one of the most expensive possible cases of such vehicles–vehicles designed to interface with a manned space station.
This is a typical tactic of opponents of an idea — to pick a worst case, and sort of imply that it’s a best one. It is not a realistic assumption, and was one of the flaws that jumped out at me at the time.
Another one is the concern about departure windows. Yes, launching from earth gives you more flexibility, but so what? If there are multiple opportunities per month from orbit (and there are, depending on how much you want to pay in delta vee for wider windows), then it doesn’t help you much to be able to launch more often from earth when your HLV architecture won’t be able to afford to launch more than a few times a year, even using the cost estimates of its most ardent proponents.
[Tuesday morning update]
Clark Lindsey has further thoughts.
[Bumped]
I’ll be curious to see your response. I just posted a comment linking him to Jon Goff’s stuff.
Propellant depots are a non-linear disturbance to the status quo. We can make some intelligent guesses about what changes will happen with them, but we can’t evaluate the entire set of changes that *can* happen with them.
PS: Rand, can I change my periodic yearnings for a pony to periodic yearnings for a LEO propellant depot? I might have a better shot at getting the latter.
can I change my periodic yearnings for a pony to periodic yearnings for a LEO propellant depot?
I wouldn’t know how go about stopping you. Or even reliably detect such a change… 😉
Tom,
Josh Hopkins is a pretty smart guy, who happens to know many of the ULA propellant depot people on a first name basis (he’s coauthored papers with many of them). His article even expresses a decent amount of familiarity with many of the counterarguments. If he hadn’t put the strawman in about trying to show that depot are 4x more expensive than CxP per mission, it would’ve actually been a pretty good piece. Now that depots are being seriously investigated, hopefully it will be possible to flesh-out more details.
My one concern has been that with my limited time resources (I’m not paid to do depot stuff), it’s hard to put together a bullet-proof and fully detailed CONOPS that you could use to “prove” that depots would be cheaper. It’s not that I haven’t written about almost every issue he’s raised, it’s just that it’s hard to provide that level of detail without actual money and large amounts of work time. A couple million bucks to do a CE&R type exercise using all the latest thinking would be really interesting.
~Jon
Thinking beyond LEO, do we have the technology to employ uranium as a propulsion power source? There’s an untapped source in a nearby one-sixth-gee environment.
http://www.space.com/scienceastronomy/090629-uranium-moon.html
A couple million bucks to do a CE&R type exercise using all the latest thinking would be really interesting.
A couple million bucks?! Are you crazy?!!!
We can’t afford that! We have heavy-lift launch vehicles to develop!
How much was the whole CE&R effort? Something like $10-20M for all the different teams that put stuff in?
Just reminds me of the old joke about “we can’t afford to do it right the first time, but we should have enough to do it again….”
~Jon
“Thinking beyond LEO, do we have the technology to employ uranium as a propulsion power source? There’s an untapped source in a nearby one-sixth-gee environment.”
Requiring a major infrastructure investment on the Moon to mine and bring up to the necessary degree of enrichment (unless you plan to send the ore to Earth, which has its own set of problems), as I pointed out in the Space.com thread (as delphinus100)…
Something like $10-20M for all the different teams that put stuff in?
Something like that.
Mining uranium on the moon makes no sense in any relevant planning horizon.
Mining uranium on the moon makes no sense in any relevant planning horizon.
I agree. The only benefit of it is launch safety concerns, and that’s not sufficient to justify the cost of ISRU development for a substance of such little mass. ISRU only makes sense in the context of bulk materials that would be very expensive to launch all the way from earth.
Considering launch windows: staging from L1 and/or SEL-2, which is an excellent idea anyway, also solves some of those problems. At least the nodal regression issue.
The issue with propellant tankers is that if you have a dumb tanker, you must have a smart depot. The tanker has to have at least enough smarts to do the approach to the depot without whacking it. Berthing the thing then does some simplification, but not that much.
You still have to be three axis stabilized to at least 1 degree, which drives a lot of hardware cost. You have to have some delta v for it, or upgrade the upper stage of the delivery vehicle, which is 6 of one and half dozen of the other.
I would probably opt for a semi smart tanker and a smart depot. Make all of the prox ops hardware common and then remove it all from the tanker after it is empty. Then you can start stockpiling parts that other buyers can use as well.
The issue with propellant tankers is that if you have a dumb tanker, you must have a smart depot. The tanker has to have at least enough smarts to do the approach to the depot without whacking it. Berthing the thing then does some simplification, but not that much.
What about tugs? I thought the idea behind a space tug was to allow for mating dumb tankers and depots?
“What about tugs? I thought the idea behind a space tug was to allow for mating dumb tankers and depots?”
Exactly. What would it take to turn an ATV into a dedicated Tug?
There’s a paper on the ULA website titled “System-of-Space Systems Architecture Utilizing Existing Space Assets to Complete and Re-Supply the International Space Station” that describes this. An existing ATV or HTV could be used as an interim tug with fairly minor modifications. It could be used two or three times in tug mode before it runs out of fuel.
If it’s refuelable it could be used far more often. With in-flight refueling you would not need to pay for development of a separate depot straight away. A dedicated tug based on the SM and without the Integrated Cargo Carrier would be more efficient. You might not recover the development costs on ISS resupply only, but in the longer run it would enable efficient resupply of both space stations and depots, which is a strategically important milestone.
I’m thinking a bit too far ahead, making use of an energy source (the Moon) that’s on the way to where we eventually want to go.
But we need an efficient means of getting off Earth first. The idea of using chemical propulsion to lift chemical propellant into orbit sounds awfully inefficient. It may be useful for some short term projects, but in the long term we need some huge innovation in launch technology.
Two inspirations I know about are space elevator and mass drivers. I don’t know how far off launch applications of either technology are, or how long construction would take. (Or the state of US relations with high-altitude equatorial nations, regarding the latter project.)
The Moon will be a practical fuel source only when and if it can launch using local resources.
What about tugs? I thought the idea behind a space tug was to allow for mating dumb tankers and depots?
You still have to have some nominal system to keep the tanker in some sort of good shape. A tug would help though.
I got this article… and agreed with it… until he got to comparing the COTS costs of shipping stuff to the ISS to what it might cost to ship up propellant into orbit.
Shipping up propellant is a relatively straight-forward task: You build the rocket in some way that has the necessary avionics (not a trivial thing, and I get that), and whatever propellant is left over can be transferred over to the fuel depot. The whole thing becomes a giant fuel tank, and worries about man-rating the vehicle can legitimately go out the window. The only real failure mode to worry about is bumping into the fuel depot. BTW, the Shuttle does a fuel transfer prior to jettisoning the external tank, bringing in fuel to “top off” what is kept on board the orbiter…. and fuel is often left in the external tank as well in not insignificant quantities either.
For me, the “container” for shipping up fuel is a no-brainer… just transfer over the reserve fuel and leave enough in the delivery vehicle to deorbit. The rest is engineering, not physics. I don’t get where the cost for delivering fuel is the same as delivering a pressurized, man-rated vehicle to the ISS and bulk cargo comprised of potentially biological experiments. It isn’t the same thing, and it certainly shouldn’t cost the same. I would certainly put it at closer to the $5k/lb. than the $30k/lb. figure quoted, and it could in theory be quite a bit cheaper. Perhaps cheap enough to kill off any ISRU potential from the Moon… at least for things leaving directly from LEO.
Rand, Jon, and others.
Thanks for the comments. For what it’s worth, I have responded to Jon’s original feedback at The Space Review website.
Josh Hopkins
The other nice thing about fuel is it is near-infinitely divisible.
You could use a laser launch, a rail gun or a Gerald Bull type space cannon(or all of the above and more) to send it up in chunks of a few hundred pounds as long as your prox ops vehicles could catch them and deal with them.
Gerald Bull type space cannons are the sort of thing that get you killed under suspicious circumstances…
Only if you point them at Israel.
This is a typical tactic of opponents of an idea — to pick a worst case, and sort of imply that it’s a best one. It is not a realistic assumption, and was one of the flaws that jumped out at me at the time.
And of course, it assumes no one will ever invent a reusable vehicle that can deliver propellant.
Even if analysis showed that heavy lift could beat propellant depots at this year’s prices, that wouldn’t necessarily mean that heavy lift was a better investment. An investor needs to consider possible future prices.
The HLV architecture is an end point — Mike Griffin touted Ares as “the launch system America will use for the next 40 years.” It will never get very much cheaper. The propellent depot architecture is a *starting* point.
A lot of people in the computer industry bought heavy lifters (mainframes) because they seemed to be cheaper than microcomputers at the time. A few years later, many (not all) of them regretted their investment. No one was foolish enough to invest in a mainframe designed to last 40 years, of course.
Some straw man objections to depots I’ve heard recently together with counterarguments:
– depots only make sense for high mission rates (no they don’t, they are definitely better than HLV at low mission rates and with RLVs may be much better at high rates)
– depots need to be so large you need an HLV to launch them (not with L1 rendez-vous)
– Congress will not pay for depots unless there is a destination first (should apply equally to HLV and whichever gets built first makes it likely the other alternative will not be built)
– propellant transfer has never been demonstrated beyond LEO (why does that matter?)
– depots are a nice add-on, HLV is crucial (it’s the other way around at best)
– depots should be left to the private sector (true for development and operation, not true for development cost, unless NASA abandons exploration plans)
Straw man isn’t the right word for all of these objections. Some of them are just plainly wrong.
– depots only make sense for high mission rates (no they don’t, they are definitely better than HLV at low mission rates and with RLVs may be much better at high rates)
Depends on how you define “low.” You could certainly define a mission rate low enough that depots wouldn’t make sense.
The question is, why would anyone want low mission rates? If spaceflight is useful and valuable, we should want to do it frequently. If it isn’t, we shouldn’t want to do it at all.
“All the universe or nothing, which shall it be?”
– Congress will not pay for depots unless there is a destination first
That one might be true, but it’s an easily testable proposition. All the Administration needs to do is go to Congress and ask.
Generally speaking, Congress funds whatever NASA and OMB ask for, give or take <5%. I have no reason to think propellant depots would be an exception. They might not be fully funded at the level the Administration requests, but I doubt they would be rejected out of hand.
For very low mission rates HLV fixed costs would dominate and a small depot might be more economical, even at current launch prices. Especially if you use in-flight refueling initially, instead of a full depot.